The diagnostic yield of pediatric polysomnography based on the professional background of referring physicians

Umakanth Khatwa*, Sriram Ramgopal, Kanwaljit Singh, Tobias Loddenkemper, Marcin Zarowski, Sanjeev V. Kothare

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

Introduction. Polysomnography (PSG) is the gold standard in diagnosis of sleep-disordered breathing but is labor intensive and expensive. We assessed PSG outcomes based on physician specialty and whether prior referral to sleep clinics reduces costs for children with sleep problems. Methods. Retrospective analysis of PSGs was performed by demographics, physician specialization, PSG indication, and results. Average costs of evaluating PSGs ordered by primary care physicians (PCPs) versus initial referral to sleep specialists were compared. Results. A total of 493 PSGs were performed between June 2010 and December 2010. Most referrals came from sleep physicians (31%), otolaryngologists (24%), and PCPs (18%). Forty-four percent PSGs ordered by PCPs were normal versus 22% by sleep physicians (odds ratio = 2.8, P <.05). Average cost of PSGs ordered by PCPs was $3053.54/patient versus $917.85/patient for sleep specialist referral. Conclusion. PSGs ordered by PCPs were more likely to be normal than PSGs by sleep physicians. Sleep clinic assessment before PSGs may assist in appropriate service utilization, improve delivery of care, and reduce health care costs by using these tests appropriately.

Original languageEnglish (US)
Pages (from-to)1154-1160
Number of pages7
JournalClinical pediatrics
Volume52
Issue number12
DOIs
StatePublished - Dec 1 2013

Keywords

  • children
  • physician sub-specialty
  • sleep studies

ASJC Scopus subject areas

  • Pediatrics, Perinatology, and Child Health

Fingerprint Dive into the research topics of 'The diagnostic yield of pediatric polysomnography based on the professional background of referring physicians'. Together they form a unique fingerprint.

Cite this